Advanced

Value at Risk - The Square Root Rule

Elsner, Yair (2012) NEKN01 20121
Department of Economics
Abstract (Swedish)
This paper tests the “Square Root Rule” (the SRR), a Basel sanctioned method of scaling 1-day Value At risk to higher time horizons. The SRR has come under serious assault from leading researchers focusing on its week theoretical basis: assuming i.i.d. asset returns. I performed an empirical test of the SRR on the 10-day horizon, the maximum horizon allowed by Basel, comparing SRR performance to directly estimating 10-day VaR by the same method 1-day was estimated, the “main-model”. This test was performed for a number of well-known models. Performance was measured by means of Basel’s own criteria for back-testing, the Christoffersen test and through general descriptive statistics. My results were that the SRR performs less well than its... (More)
This paper tests the “Square Root Rule” (the SRR), a Basel sanctioned method of scaling 1-day Value At risk to higher time horizons. The SRR has come under serious assault from leading researchers focusing on its week theoretical basis: assuming i.i.d. asset returns. I performed an empirical test of the SRR on the 10-day horizon, the maximum horizon allowed by Basel, comparing SRR performance to directly estimating 10-day VaR by the same method 1-day was estimated, the “main-model”. This test was performed for a number of well-known models. Performance was measured by means of Basel’s own criteria for back-testing, the Christoffersen test and through general descriptive statistics. My results were that the SRR performs less well than its “main-model” alternative, but overall only slightly so. Based on this I claim that SRR usage as a thumb-rule is certainly defensible, and that even in circumstances calling for exact measures it’s difficult to rule out SRR-scaling. (Less)
Please use this url to cite or link to this publication:
author
Elsner, Yair
supervisor
organization
course
NEKN01 20121
year
type
H1 - Master's Degree (One Year)
subject
keywords
Value At Risk, VaR, scaling, risk, risk-measurement, financial risk, downside-risk, Basel
language
English
id
2426134
date added to LUP
2012-04-04 13:43:06
date last changed
2012-04-04 13:43:06
@misc{2426134,
  abstract     = {This paper tests the “Square Root Rule” (the SRR), a Basel sanctioned method of scaling 1-day Value At risk to higher time horizons. The SRR has come under serious assault from leading researchers focusing on its week theoretical basis: assuming i.i.d. asset returns. I performed an empirical test of the SRR on the 10-day horizon, the maximum horizon allowed by Basel, comparing SRR performance to directly estimating 10-day VaR by the same method 1-day was estimated, the “main-model”. This test was performed for a number of well-known models. Performance was measured by means of Basel’s own criteria for back-testing, the Christoffersen test and through general descriptive statistics. My results were that the SRR performs less well than its “main-model” alternative, but overall only slightly so. Based on this I claim that SRR usage as a thumb-rule is certainly defensible, and that even in circumstances calling for exact measures it’s difficult to rule out SRR-scaling.},
  author       = {Elsner, Yair},
  keyword      = {Value At Risk,VaR,scaling,risk,risk-measurement,financial risk,downside-risk,Basel},
  language     = {eng},
  note         = {Student Paper},
  title        = {Value at Risk - The Square Root Rule},
  year         = {2012},
}